AI News

\texttt{Range-Arithmetic}: Verifiable Deep Learning Inference on an Untrusted Party (arxiv.org)

TileFuse: A Fused Mixed-Precision Kernel Library for Efficient Quantized LLM Inference on AMD NPUs (arxiv.org)

When Poison Fails After Retrieval: Revisiting Corpus Poisoning under Chunking and Reranking Pipelines (arxiv.org)

OmniBioTwin: A System-of-Twinned-Systems Framework for Health Digital Twins (arxiv.org)

RAIL: Rethinking Auditory Intelligence in Large Audio-Language Models with a CHC-Grounded Benchmark (arxiv.org)

SPEAR: A System for Post-Quantization Error-Adaptive Recovery Enabling Efficient Low-Bit LLM Serving (arxiv.org)

Artificial Intelligence in Ship Finance: Applications, Opportunities, and a Case Study in AI-Augmented Loan Origination (arxiv.org)

Learning to Inject: Automated Prompt Injection via Reinforcement Learning (arxiv.org)

Vector Quantized Latent Concepts: A Scalable Alternative to Clustering-Based Concept Discovery (arxiv.org)

AI4SLT: Empirical Processes in Lean 4 for Formal Statistical Learning Theory (arxiv.org)

Prediction-Powered Risk Monitoring of Deployed Models for Detecting Harmful Distribution Shifts (arxiv.org)

Open Materials Generation with Inference-Time Reinforcement Learning (arxiv.org)

Towards a Bridge Layer Between Bibliographic and Formalized Mathematical Knowledge (arxiv.org)

Afrispeech Semantics: Evaluating Audio Semantic Reasoning in Spoken Language Models Across Domains and Accents (arxiv.org)

An Ethical eValuation Agent (EeVA): Results of a Proof-of-Concept Test on a Prototype Agentic-like Workflow to Assist Ethical Deliberations (arxiv.org)

Preregistration for Experiments with AI Agents (arxiv.org)

The Environmental Cost of LLMs in AIED: Reporting and Practices (arxiv.org)

From Awareness to Action: Understanding and Overcoming the Research-Practice Gap in Algorithmic Fairness for Public Health (arxiv.org)

FOCUS: DLLMs Know How to Tame Their Compute Bound (arxiv.org)

Robust Privacy: Inference-Stage Privacy through Certified Robustness (arxiv.org)

Robustness of Mixtures of Experts to Feature Noise (arxiv.org)

Analytic Bijections for Smooth and Interpretable Normalizing Flows (arxiv.org)

Learning Object Manipulation from Scratch via Contrastive Interaction (arxiv.org)

Every Act Has Its Price: Compressed Moral Composition in Frontier LLMs (arxiv.org)

T2MM: An LLM Supported Architecture For Inquiry-Based Modeling (arxiv.org)

BioDivergence: A Benchmark and Evaluation Framework for Hidden Contextual Contradictions in Biomedical Abstracts (arxiv.org)

What Limits Does Quantization Place on Dense Top-$k$ Retrieval? A Theoretical Study (arxiv.org)

Unifying Learning Dynamics and Generalization in Transformers Scaling Law (arxiv.org)

Improving Generalization and Data Efficiency with Diffusion in Offline Multi-agent RL (arxiv.org)

Towards Deep Learning Surrogate for the Forward Problem in Electrocardiology: A Scalable Alternative to Physics-Based Models (arxiv.org)

Semantic search for 100M+ galaxy images using AI-generated captions (arxiv.org)

Irresponsible AI: big tech's influence on AI research and associated impacts (arxiv.org)

Grounding Computer Use Agents on Human Demonstrations (arxiv.org)

FlexiBrain: Resolution-Agnostic Voxel-Level Encoding for Native fMRI (arxiv.org)

Parameter-Efficient Adapter Tuning for Tabular-Image Multimodal Learning (arxiv.org)

Evaluating and Combating the Impact of Concept Drift on the Performance of Machine Learning-Based Phishing Detection Systems (arxiv.org)

Density estimation for Hellinger via minimum-distance estimators: mixtures of Gaussians, log-concave, and more (arxiv.org)

APEX: Automated Prompt Engineering eXpert with Dynamic Data Selection (arxiv.org)

Moving Beyond Diffusion: Hierarchy-to-Hierarchy Autoregression for fMRI-to-Image Reconstruction (arxiv.org)

Certifiable Safe RLHF: Semantic Grounding and Fixed Penalty Constraint Optimization for Safer LLM Alignment (arxiv.org)

Noise-Guided Transport for Imitation Learning (arxiv.org)

Geometric Metrics and LLMs: What They Measure and When They Work (arxiv.org)

Toward Preference-aligned Large Language Models via Residual-based Model Steering (arxiv.org)

Spatially Masked Regression Reveals Local and Distributed Predictability in Electrophysiological Recordings (arxiv.org)

A Scalable PyTorch Abstraction for Multi-GPU Gaussian Splatting (arxiv.org)

Small Experiments, Cheaper Decisions: A Case Study in Staged Promotion for Micro-Pretraining (arxiv.org)

When Probing Accuracy Saturates, Fragility Resolves: A Complementary Metric for LLM Pre-Training Analysis (arxiv.org)

Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models (arxiv.org)

Rule Taxonomy and Evolution in AI IDEs: A Mining and Survey Study (arxiv.org)

Adapting Prithvi-EO for Fallow Detection for Food-Water Nexus: ViT-Adapter Necks and Parameter-Efficient Backbone tuning of Geospatial Foundation Model (arxiv.org)